Sensor location determination for geographical region

1. A device for determining where to situate sensors in a geographical region to account for attribute variation of soil or foliage in the geographical region, the device comprising: first processing circuitry configured to: receive image data of an image of the geographical region, the image data representative of an attribute of the soil or foliage; receive soil data regarding physical characteristics of soil of cells within the geographical region; generate a feature vector based on the image data and the soil data; and produce a matrix including entries indicating how similar the cells are in terms of the image data and the soil data; second processing circuitry to implement a clustering circuitry configured to: receive the matrix as input; and produce data indicating a cluster of clusters to which each cell of the cells belongs, each cell more similar, in terms of the soil data and image data, to other cells of the cluster to which they belong than cells of other clusters, each cluster indicating a location at which to situate a sensor of the sensors to monitor the attribute.

2. The device of claim 1 , wherein the image data includes one of multi-spectral image data and color image data, and the attribute includes one of soil moisture and soil temperature.

3. The device of claim 1 , wherein the physical characteristics of the soil include at least one of soil texture, soil structure, consistence, particle density, bulk density, pore space, color, and permeability.

4. The device of claim 1 , wherein the feature vector includes, for each cell of the cells, an average of the image data corresponding to the cell and soil data corresponding to the cell for one or more of the physical characteristics.

5. The device of claim 4 , wherein the feature vector includes a row of data for each cell of the cells, each row including data for that cell.

6. The device of claim 1 , wherein producing the matrix includes using a radial basis function with the feature vector as input.

7. The device of claim 6 , wherein the radial basis function is a Gaussian radial basis function.

8. The device of claim 1 , wherein producing the data indicating a cluster of clusters to which each cell of the cells belongs includes using a k-means clustering unsupervised machine learning technique.

9. The device of claim 1 , wherein the image data is first image data of a first image and is representative of a first attribute, the feature vector is a first feature vector, the matrix is a first matrix, the data is first data, the cluster and clusters are a first cluster and first clusters, respectively, and the sensor and sensors are a first sensor and first sensors, respectively, and wherein the first processing circuitry is further configured to: receive second image data of a second image of the geographical region, the second image data representative of a second attribute of the soil or foliage; generate a second feature vector based on the second image data and the soil data; and produce a second matrix including entries indicating how similar the cells are in terms of the image data and the soil data; the second processing circuitry is further configured to: receive the second matrix as input; and produce second data indicating a second cluster of second clusters to which each cell of the cells belongs, each cell more similar, in terms of the soil data and image data, to other cells of the second cluster to which they belong than cells of other second clusters, each second cluster indicating a location at which to situate a second sensor of second sensors to monitor the second attribute.

10. The device of claim 9 , wherein the first processing circuitry is further configured to: compare the first clusters and the second clusters to determine locations in the geographical region where a cluster of the first cluster overlaps with a cluster of the second clusters; and produce data indicating the locations of overlap in which to place both (a) a sensor of the first sensors and (b) a second sensor of the second sensors.

11. A method for determining where to situate sensors in a geographical region to account for attribute variation of one or more attributes of soil or foliage in the geographical region, the method comprising: generating, by processing circuitry, a feature vector based on image data and soil data, the image data of an image of the geographical region produced by an aerial vehicle, the image data representative of an attribute of the attributes of the soil or foliage, and the soil data indicating physical characteristics of soil of cells within the geographical region; producing a matrix including entries indicating how similar the cells are in terms of the image data and the soil data based on the feature vector; and producing, at clustering circuitry and based on the matrix, data indicating a cluster of clusters to which each cell of the cells belongs, each cell more similar, in terms of the soil data and image data, to other cells of the cluster to which they belong than cells of other clusters, each cluster indicating a location at which to situate a sensor of the sensors to monitor the attribute.

12. The method of claim 11 , wherein the image data includes one of multi-spectral image data and color image data, the attribute includes one of soil moisture and soil temperature, and the physical characteristics of the soil include at least one of soil texture, soil structure, consistence, particle density, bulk density, pore space, color, and permeability.

13. The method of claim 11 , wherein the feature vector includes, for each cell of the cells, an average of the image data corresponding to the cell and soil data corresponding to the cell for one or more of the physical characteristics.

14. The method of claim 11 , wherein the image data is first image data of a first image and is representative of a first attribute, the feature vector is a first feature vector, the matrix is a first matrix, the data is first data, the cluster and clusters are a first cluster and first clusters, respectively, and the sensor and sensors are a first sensor and first sensors, respectively, and wherein the method further comprises: receiving second image data of a second image of the geographical region, the second image data representative of a second attribute of the soil or foliage; generating a second feature vector based on the second image data and the soil data; producing a second matrix including entries indicating how similar the cells are in terms of the image data and the soil data; and producing second data indicating a second cluster of second clusters to which each cell of the cells belongs, each cell more similar, in terms of the soil data and image data, to other cells of the second cluster to which they belong than cells of other second clusters, each second cluster indicating a location at which to situate a second sensor of second sensors to monitor the second attribute.

15. The method of claim 14 , further comprising: comparing the first clusters and the second clusters to determine locations in the geographical region where a cluster of the first cluster overlaps with a cluster of the second clusters; and producing data indicating the locations of overlap in which to place both (a) a sensor of the first sensors and (b) a second sensor of the second sensors.

16. A non-transitory machine-readable medium including instructions that, when executed by a machine, configure the machine to perform operations comprising: generating a feature vector based on image data and soil data, the image data of an image of the geographical region produced by an aerial vehicle, the image data representative of an attribute of attributes of the soil or foliage, and the soil data indicating physical characteristics of soil of cells within the geographical region; producing a matrix including entries indicating how similar the cells are in terms of the image data and the soil data based on the feature vector; and producing data indicating a cluster of clusters to which each cell of the cells belongs, each cell more similar, in terms of the soil data and image data, to other cells of the cluster to which they belong than cells of other clusters, each cluster indicating a location at which to situate a sensor of sensors to monitor the attribute.

17. The non-transitory machine-readable medium of claim 16 , wherein the feature vector includes a row of data for each cell of the cells, each row including data for that cell, and wherein producing the data indicating a cluster of clusters to which each cell of the cells belongs includes using a k-means clustering unsupervised machine learning technique.

18. The non-transitory machine-readable medium of claim 16 , wherein producing the matrix includes using a radial basis function to generate the matrix with the feature vector as input, and wherein the radial basis function is a Gaussian radial basis function.

19. The non-transitory machine-readable medium of claim 16 , wherein the image data is first image data of a first image and is representative of a first attribute, the feature vector is a first feature vector, the matrix is a first matrix, the data is first data, the cluster and clusters are a first cluster and first clusters, respectively, and the sensor and sensors are a first sensor and first sensors, respectively, and wherein the operations further comprise: receiving second image data of a second image of the geographical region, the second image data representative of a second attribute of the soil or foliage; generating a second feature vector based on the second image data and the soil data; producing a second matrix including entries indicating how similar the cells are in terms of the image data and the soil data; and producing second data indicating a second cluster of second clusters to which each cell of the cells belongs, each cell more similar, in terms of the soil data and image data, to other cells of the second cluster to which they belong than cells of other second clusters, each second cluster indicating a location at which to situate a second sensor of second sensors to monitor the second attribute.

20. The non-transitory machine-readable medium of claim 19 , wherein the operations further comprise: comparing the first clusters and the second clusters to determine locations in the geographical region where a cluster of the first cluster overlaps with a cluster of the second clusters; and producing data indicating the locations of overlap in which to place both (a) a sensor of the first sensors and (b) a second sensor of the second sensors.